Table 1.
Comparison of the partial AUC of competing regulatory network learning algorithms for the first simulated network. mean.pAUC.1 and mean.pAUC.2 are the average of partial AUC over 30 target mRNAs/proteins for the first and the second algorithm, respectively
| Alg1 | Alg2 | Sample size | Target | mean.pAUC.1 | mean.pAUC.2 | wilcox. pval |
|---|---|---|---|---|---|---|
| MBSmRN | MMHC | 1250 | mRNA | 0.9226 | 0.8487 | 0.01463 |
| MBSmRN | Lasso | 1250 | mRNA | 0.9226 | 0.6130 | 8.41E-09 |
| MBSmRN | Lasso (no protein)a | 1250 | mRNA | 0.9226 | 0.5328 | 2.38E-11 |
| MMHC | Lasso | 1250 | mRNA | 0.8487 | 0.6130 | 1.29E-07 |
| MBSmRN | MMHC | 250 | mRNA | 0.9068 | 0.8228 | 0.005421 |
| MBSmRN | Lasso | 250 | mRNA | 0.9068 | 0.8169 | 0.009452 |
| MBSmRN | Lasso (no protein) | 250 | mRNA | 0.9068 | 0.6112 | 2.42E-09 |
| MMHC | Lasso | 250 | mRNA | 0.8228 | 0.8169 | 0.80688 |
| MBSmRN | MMHC | 50 | mRNA | 0.8182 | 0.6849 | 0.000688 |
| MBSmRN | Lasso | 50 | mRNA | 0.8182 | 0.8502 | 0.41131 |
| MBSmRN | Lasso (no protein) | 50 | mRNA | 0.8182 | 0.6388 | 3.17E-06 |
| MMHC | Lasso | 50 | mRNA | 0.6849 | 0.8502 | 8.77E-06 |
| MBSNIG | MMHC | 1250 | Protein | 0.9970 | 0.7391 | 6.09E-10 |
| MBSNIG | Lasso | 1250 | Protein | 0.9970 | 0.5905 | 2.64E-12 |
| MBSNIG | Lasso (no protein) | 1250 | Protein | 0.9970 | 0.5070 | 2.35E-12 |
| MMHC | Lasso | 1250 | Protein | 0.7391 | 0.5905 | 5.59E-05 |
| MBSNIG | MMHC | 250 | Protein | 0.9967 | 0.7067 | 4.10E-11 |
| MBSNIG | Lasso | 250 | Protein | 0.9967 | 0.7460 | 1.12E-11 |
| MBSNIG | Lasso (no protein) | 250 | Protein | 0.9967 | 0.5029 | 1.52E-12 |
| MMHC | Lasso | 250 | Protein | 0.7067 | 0.7460 | 0.420141 |
| MBSNIG | MMHC | 50 | Protein | 0.9372 | 0.6214 | 7.56E-09 |
| MBSNIG | Lasso | 50 | Protein | 0.9372 | 0.8257 | 7.04E-05 |
| MBSNIG | Lasso (no protein) | 50 | Protein | 0.9372 | 0.5080 | 2.90E-11 |
| MMHC | Lasso | 50 | Protein | 0.6214 | 0.8257 | 1.05E-06 |
aLasso algorithm is applied to simulation data with the protein data removed.